Affinities of Pteronarcys regalis. 437 



glia, which are somewhat larger than corresponding tracheae in other insects 

 in which the lateral abdominal trunks are dilated into sacs. A curious ter- 

 mination of the lateral abdominal trunks occurs in the posterior segments. 

 These trunks end abruptly immediately after their junction (fig. 11) in a kind 

 of caecum, from which a small branch extends backwards to the caudal styles, 

 analogous to the mode in which the branchial filaments are supplied from the 

 main tracheae. One of the most curious distributions of tracheae in Ptero- 

 narcys is of those which are supplied to the alimentary canal. Tracheae which 

 supply this organ are rarely or ever dilated in any insect, not even in theHy- 

 menoptera, in which the sacs of the main trunks are the largest. They pass 

 off as slender branches, either from the lateral sacs or from the main trunks 

 in the immediate vicinity of the spiracles, and decrease in size as they are 

 distributed on the canal, as I have figured and described* in the male of 

 Bombus terrestris. A slender branch passes longitudinally backwards in that 

 insect from the main trachea, behind the metathoracic spiracle on each side 

 of the oesophagus, to the anterior portion of the stomach on which it is distri- 

 buted, and a similar origin and distribution of the gastric tracheae exists in 

 all insects with but little variation. In Pteronarcys a long, slender branch 

 (fig. 10 i, k) passes oflf from the slightly enlarged trunk behind the mesotho- 

 racic spiracle, and another [l) from behind the metathoracic, which are ex- 

 tended longitudinally backwards, slightly reduced in size, as far as the middle 

 of the abdomen, where they divide into branches which are distributed on the 

 sides and anterior of the stomach. This is the general character of the tra- 

 cheae in the whole of the Perlidce and in Sialis. These exceptions to the law 

 which I have heretofore endeavoured to exemplify by facts, that a vesicular 

 form of the respiratory organs in insects has reference chiefly to power of 

 flight, and enables the insect to alter the specific gravity of its body at the 

 moment it takes wing, and thus diminish the amount of muscular exertion 

 required in its movements, tend in reality to confirm the previous conclu- 

 sions. The retention in the imago of the simple setaceous tracheae of the 

 larva is accompanied, as in Sialis, Perla, and Pteronarcys, with a low power 

 of flight, although the species of each of these genera have ample wings, and 

 might have been expected to be extremely active. Pteronarcys thus resembles 



* Philosophical Transactions, 1836, part ii. pi. 26. fig. 2. p. 564. 



